It is generally desirable to be able sensitively, specifically, qualitatively and/or quantitatively to detect RNA, and in particular mRNA, in a sample, including for example in fixed or fresh cells or tissues. It may be particularly desirable to detect an mRNA in a single cell. For example, in population-based assays that analyze the content of many cells, molecules in rare cells may escape detection. Furthermore, such assays provide no information concerning which of the molecules detected originate from which cells. Expression in single cells can vary substantially from the mean expression detected in a heterogeneous cell population. It is also desirable that single-cell studies may be performed with single-molecule sensitivity which allows the fluctuation and sequence variation in expressed transcripts to be studied. Fluorescence in situ hybridization (FISH) has previously been used to detect single mRNA molecules in situ. Although permitting determination of transcript copy numbers in individual cells, this technique cannot resolve highly similar sequences, so it cannot be used to study, for example, allelic inactivation or splice variation and cannot distinguish among gene family members.
The only option available for assigning transcript variants to a single cell in a given tissue involves polymerase chain reaction (PCR) of laser-capture microdissected material, which is time consuming and error prone, and thus not suitable for diagnostics.
As an alternative to PCR- and hybridization-based methods, padlock probes (Nilsson et al., 1994) have for many years been used to analyze nucleic acids. These highly selective probes are converted into circular molecules by target-dependent ligation upon hybridization to the target sequence. Circularized padlock probes can be amplified the localization of target molecules, including, where DNA targets are concerned, at the single-cell level. Such a protocol is described in Larrson et al., 2004), in which the target DNA molecule is used to prime the RCA reaction, causing the RCP to be anchored to the target molecule, thereby preserving its localization and improving the in situ detection.
While RNA molecules can also serve as templates for the ligation of padlock probes (Nilsson et al., 2000), RNA detection with padlock probes in situ has so far proven more difficult than DNA detection and is subject to limitations (Lagunavicius et al., 2009). For example, the high selectivity reported for padlock probes with in situ DNA detection and genotyping has not been reproduced with detection of RNA targets in situ. This is possibly due to problems with ligation of DNA molecules on an RNA template, since it is known that both the efficiency and the specificity of the ligation reaction are lower compared to ligation on a DNA template (Nilsson et al., 2000; Nilsson et al. 2001). It has recently been demonstrated that RNA molecules may be detected in situ with padlock probes and target-primed RCA (Lagunavicius et al., 2009; Stougaard et al., 2007). However, thus far, detection through target-primed RCA has for the most part been restricted to sequences in the 3′-end of non-polyadenylated RNA or sequences adjacent to the poly(A)-tail of mRNA. Since target-priming of the RCA reaction is dependent on a nearby free 3′-end that can be converted into an RCA primer, it is thought that this limitation results from the formation of RNA secondary structures which impede the polymerase action (3′ exonucleolysis) required to convert the RNA into a reaction primer. The detection efficiency of direct mRNA detection with padlock probes has been estimated to be as low as 1% (Nilsson et al., 2001). For the detection of non-polyadenylated RNA molecules, it has been noted that ligation of the probes using an internal hairpin structure as template resulted in higher detection efficiency than using the RNA molecule itself as ligation template (Stougaard et al., 2007). This indicates that better ligation conditions are required to be able to efficiently detect and genotype RNA directly with padlock probes in situ.
None of the methods for in situ detection of RNA presented thus far provide the possibility to detect sequence variation at the single nucleotide level and in particular to genotype transcripts. In the present embodiment, by converting an RNA target molecule into cDNA, the reduction in padlock probe ligation efficiency and specificity is avoided and the excellent genotyping properties provided by padlock probes are preserved. In addition, it has been found that unlike many previously described methods, embodiments are not restricted to detection of sequences positioned at specific sites in the RNA molecules.